Measuring metabolic fluxes in living cells by 13C metabolic flux analysis has become a key technology for improving our quantitative understanding of cellular metabolism. After two and a half decades of development, driven by diverse analytical and computational innovations, a rich set of tools has become available supporting all steps of the 13C MFA workflow, ranging from the assembly of high-fidelity models and their efficient simulation to convenient design of informative tracer compositions. For studying complex biological systems in less defined environments, such as pathogens residing in a host, however, many challenges remain. One critical step is the identification of a “useful” model formulation with which the fluxes are to be inferred from the data at hand. In the talk, I will present new directions in 13C MFA, powered by Bayesian statistics, to go about this hitherto neglected question and show first successful applications for Escherichia coli and Mycobacterium tuberculosis. Following this path keeps promise to strengthen the position of 13C MFA as an epistemic tool for explaining phenotypes and building models of metabolism.